DOCUMENT RESUME
ED 276 257 FL 016 209
AUTHOR Peyton, Joy EreeftTITLE Dialogue Journal Writing and the Acquisition of
English Grammatical Morphology.INSTITUTION Center for Applied Linguistics, Washington, D.C.SPONS AGENCY National Inst. of Education (ED), Washington, DC.PUB DATE [86]GRANT NIE-G-83-0030NOTE 52p.; For related documents, see FL 016 208 and FL
016 139-141.PUB TYPE Reports - Research/Technical (143)
EDRS PRICE MF01/PC03 Plus Postage.DESCRIPTORS Classroom Techniques; Dialogs (Language); *English
(Second Language); *Grammar; InstructionalEffectiveness; Interlanguage; *Learning Processes;*Morphology (Languages); Second Language Instruction;Second Language Learning; Transfer of Training
IDENTIFIERS *Dialogue Journals
ABSTRACTA study of the language used in the dialogue journals
of beginning students of English as a second language (ESL) focusedon the acquisition of English morphology. The study used twomethodological approaches: a comparison of journal language with thatused in speech and other written samples, and a longitudinal look atchange patterns that also examined which morphemes were beingacquired and which were not. Four major findings emerged: (1) somemorphemes, including the use of "be" as a copula, were masteredquickly, while others showed little or no gain over tine; (2) whilethere were some overall trends common to all students, there was alsoconsiderable individual variation in the acquisition processes,seemingly as a result of first language transfer; (3) dialoguejournal writing does reflect changes in language proficiency overtime, even at the earliest stages of language learning; and (4) ESLlearners can compose and express themselves in English long beforethey have mastered its forms and structures. References, tables, andfigures are appended. (MSE)
********************w*************************************************** Reproductions supplied by EDRS are the best that can be made ** from tho original document. ************************************************************************
DIALOGUE JOURNAL WRITING
AND THE ACQUISITION OF
ENGLISH GRAMMATICAL MORPHOLOGY
JOY KREEFT PEYTON
U.IL DEPARTMENTOF EDUCATIONMe* of &lumen/.Reeeeten end improvement
EDUCATIONAL RESOURCESINFORMATION
CENTER (ERIC)Az. document nee been
ropfeducoa meyed from the mean or Oriplmaliononeutiltuto ItC Mum/ changes Niu
boon made to imixontoltOmduction quality
Points of vow/ oomfone Meted m the docu .min dO nOI neceSeentyteptMmt *MOMOERi moon of MALI'
"PERMISSION TO REPRODUCE THISMATERIAL HAS BEEN GRANTED BY
TO THE EDUCATIONAL RESOURCESINFORMATION CENTER (ERIC)."
Dialogue Journal Writing
and the Acquisition of English Grammatical Morphology
Joy Ereeft Peyton
During the past few years, considerable interest has developed in the use
of dialogue journals --with native English speakers, with students learning
English as a second language (ESL), and with hearing -inpaired students, whose
exposure to spoken English is linited. dialogue journal is a bound notebook
in which students write regularly (daily, if possible) to a teacher, and the
teacher writes back to each student each time they write - -not to *valuate the
writing in any way, but as an active participant in a written "conversation"
that continues throughout the entire school term. Particular characteristics of
dialogue journal writing (such as content and amount of wTiting done) vary
according to teacher objectives and types of students involved, but in many
classes students are encouraged to write as much as they want about topics of
their own choosing. This teacher-developed practice originated with Leslee
Reed, a sixth grade teacher in Los Angeles, with native English-speaking stu-
dents. She began using dialogue journals as a way to communicate regularly with
each student about what they were learning. After using them for a number of
years, she found then to be invaluable as a source of infornation about student
interests and concerns, an a place where students could write freely and openly
about topics that interest them, and as a forum for thinking together about
issues and problems confronting the students (Staton, 1980).
A few years ago, Mrs. geed began teaching in a school in Los Angeles in
which 902 of the students were nonnative English speakers. In her sixth-grade
class of 26 students, who came frou twelve countries and ten language
backgrounds, dialogue journal writing became central to her teaching. It pro-
vided a way for her to know these students and to assist them la adjusting to
their new language and culture, through consistent and supportive interaction
3
with a member of the culture. It also provided a means for individualizing
instruction in a multilingual, multicultural classrooa, with students at various
levels of English proficiency, and for comaunicating with students at their
level of proficiency.
Besides these social and cultural values of dialogue journals, they also
showed promise as a way to proaote the language acquisition of ESL students and
hence their ability to express themselves in written English. In the dialogue
journal, even students at very beginning levels of English proficiency can pro-
duce some message on paper, even if in the form of picture cr a few words or
sentences, and receive a response. More advanced students can freely write
extended text about topics of interest to them, receive consistent feedback
about their ideas, and read text written at their reading level. At the Same
time, manners of written expression and language structures are modeled.
Dialogue journals and language acquisition
Findings from two studies of the dialogue journal writing of native English
speakers (Staton, Shuy, Ireaft, and Peed, in press) and students learning ESL
(Kreeft, Shuy, Staton, Reed, and Morroy, 1984) indicate that the interaction
that occurs in the journals has the potential for prosoting second language
acquisition, because it has many of the features of conversations between
children learning a first language and a caregiver and between second language
learners and native speakers of that language in an informal context, i.e., the
"set of requirements that should be met by any activity or set of materials
timed at subconscious language acquisition" (Krashen, 1982, pp. 62-76). Krashen
(1984) discusses the importance of these conditions in the acquisition of
writing, and Staton (1984) and Kreeft Peyton (1986) have applied them to dia-
logue journal writing:
o The interaction focuses on real topics and issues introduced by and of
interest to the learner.
o The focus of the interaction is on meaning rather than on form.
-2- 4
o The language input that the learner receives from reading the teacher's
entry is comprehensible, modified roughly to the learner's level of English pro-
ficiency, and slightly beyond the learner's productive ability. (See the
variation in the teacher's language to five different students in the examples
on page 9.)
o The dialogue moves naturally from material that is familiar to the stu-
dent (e.g., past experiences) or shared with the teacher (e.g., classroom
experiences) to the less familiar (e.g., mew experiences, new ideas, and future
plans).
o The language that occurs in the journals is not grammatically sequenced
according to some pre-established plan, but rather the use of grammatical forms
and structures evolves naturally in the process of the interaction.
o Rather than overt correction of student errors, correct grammatical
forms and structures can be modeled in the course of the interaction. Genuine
requests for clarification can remedy breakdowns of communication resulting from
errors in form.
o The continuity of the dialogue provides the opportunity to receive more
input on a given topic.
o The interaction occurs in private, in a non-threatening, supportivt con-
tex t
It appears therefore that dialogue journal writing can provide a valuable
context for the acquisition of a second language in writing, in a situation
similar to the kind of interaction that occurs in speech. This raises a number
of interesting research issues. Is there in fact evidence of language acquisi-
tion of beginning ESL learners over time in this free, unmonitored writing, and
can gains in language facility be plotted, as information for teachers about the
progress of their students? If so, what are the acquisition patterns in dia-
logue journal writing, and how do they parallel patterns of acquisition in
-3-
5
speech already documented in the language acquisition literature? Are acquisi-
tion patterns in dialogue journals particular to individual students or uniform
among students?
The study reported in this chapter seeks to address these issues by docu-
menting the acquisition of English grammatical morphology as evidenced in the
journal writing of beginning ESL learners over ten mouths' time. Of course,
morphology represents only one small aspect of language acquisition, and there
are many other avenues that could be pursued in future studies of patterns of
acquisition over time in the writing. However, morphology is a good place to
start. The frequent and obligatory occurrence of morphemes in native usage
makes quantification and hence the investigation of uniformity, variability, and
change over time feasible. Also, there already exists a large body of litera-
ture on grammatical morphology of learners of English as both a first and a
second language, primarily in oral language productions, vith which to compare
results from this study of writing.
Previous studies of grammatical morphology
The body of researt-h that informs many of the analytical methods used here
is the nmorpheme studies" of the 1970's. Dulay and Kurt (1972, 1973, 1974a,
1974b) conducted the first studios of second language learners and a plethora of
studies followed, as researchers sought to discover whether there are universal
Itrocesses that guide acquisition of English as a second language, regardless of
native language background. One approach to this question was to determine
whether there was a universal and invariant order for the accurate use of
morphemes in required contexts among ESL learners from various language and edu-
cational backgrounds, at different ages, and in both spoken and written produc-
tions (cf. Kreeft, 19.4 for a more extensive review of the morpheme studies and
a discussion of the issues raised by them).
-4-
6
The accurate use of a specific morpheme was determined by looking at a con-
text in which it is required and deciding whether or not it is supplied. For
example, in the sentence "Yesterday we a to the too," past tense le required
with As, since the adverb yesterday requires past tense on the verb, but it is
not supplied. The morphoses were then ranked in relation to each other,
according to the frequency with which they were supplied in required contexts.
Despite many criticisms of the morpheme studies regarding methods of data
collection and analysis and interpretations of findings, similar orders were
found among ESL learners in nearly all of the studies where subjects were not in
a test-like situation, indicating that there are in fact certain universal pro-
cesses of second language acquisition. Rased on these findings Krashen (1977)
proposed a "natural" or universal order for morpheme acquisition for children
and adult second language learners in free speaking and writing situations,
where the focus is primarily on the message and not on fora. This order is
shown in Figure I, the boxes representing a descending order of accuracy. That
is, in required contexts, progressive plural, and copula are "acquired" or
used before the progressive auxiliary and articles. The order of morpheues
within each box le variable.
!Figure 1 about herej
Acquisition of grammatical aorphology in dialogue journals
The data
The data for this study of morphological acquisition in the context of
student-generated, relatively unmonitored writing come from the dialogue journal
writing of two classes of sixth-grade ESL learners in Los Angeles, California.
Their teacher, Leslee Reed, le the "original" dialogue journal teacher, as
described earlier. The classes are mainstream sixth-grade classes with a few
native English speakers, some students who were born in and have had all of
their schooling la the United States but who speak a language other than English
at home, and a majority of students who have come to the United States from
.5 - 7
other countries some time during their school years--anyWhere from five years to
a few months before entering this class. A regular sixth-grade curriculum is
taught, with supplementary ESL classes for new arrivals. Except for math, all
subjects are taught by the same teacher. The dialogue journal is a supplemen-
tary activity, Which students write in during their free tine throughout the
dey. Each morning, they receive their journal with the teacher's response,
written the night before.
The following exchange illustrates the nature of the interaction in one
student's journal. These entries come from the fifth month of dialogue journal
writing between the teacher and Michael, a sixth grade student from Burma, who
had been in the United States for just over a year at the time that this
interaction was written.
February 9
Michael: Mrs. Reed, you know on this week like the silly week.I don't know what happen on this week. Mrs. Reed,what did you mean about the valentines you said wehave to bring the valentines. Did we have to made thevalentines for people in our classroom? I dont knowwhat are you talking.
Teacher: Mo, we don't have to send anyone a valentine. It isjust a fun thing to do. Souetimes vs like someone butwe do not tell thee. We feel funny telling someone wereally like thee. Diving a valentine is an easy wayof doing it. If you want to give a valentine or fiftyvalentines it doesn't matter. You do what you wantabout that.
February 10,
Michael: Mrs. Reed I know what is the valentine but I don'tknow What I have to do and the valentine is we haveto glve the cards to scmeone and I have to buy thecards but I can't buy the cards that is the problem.I think I'm not going to the sofeball becuse I did notdo nothing about it end Ricardo said I could be thecatcher and he break his primise.
Teacher: Mb problem! Anyone who wants to give lialentines can.If you don't want to give valentines you don't haveto.
Talk to Ricardo again! I'm sure he forgot his promisewhen others on tho team began yelling at him.
Did you aver find your pan?
.6-
February 11
Michael: I didn't not fiad my pen. Happy Valentine! I want togive the cards to people but I can't give the cards topeople. I give the one card to the simon. I thinkLevel 10 hard de. Did you think Level 10 is hard forme? I saw the Thanksgiving Pilgrim in the book. Youknow, today morning U Chal put cards in every bag exceptTony. I know why U Chal put the cards because he putcards into every bag so that he merger him. You said weare going to do the art with the shoe box, and you saidyou don't have the shoe box you can't do art.
Teacher: I am not sure What you said. U Chul put cards ineveryone's sack so they don't merger him? What word didyou put there? Please tell me.Yes, we will do the shoebox art, iv-ryone 4ho has ashoebox will do it. Have a happy vacation. I will seeyou in five days.
This ongoing, daily writing consists of genuine and spontaneous interaction
about topics that both the teacher and the students introduce and develop over
ties. Because the interaction is written, more than one topic can be introduced
in a "turn," and topics are continued for a number of turns (e.g., the
discussion of valentines) or introduced and dropped (the discussion of Ricardo
and softball), dependirl on student and teacher interests. Both people give
information and opinions, ask and answer questions, and request explanations and
clarification.
Five students were chosen for the study, from two different classes of 26
students, during the 1980-8i and 1981-82 school years (Septenber to June).
Figure 2 gives background information on each student. Those particular stu-
dents were chosen because they were beginning ESL learners, who had been in the
United States for less than one year when they began writing in the journals. I
was interested in plotting the development of students in the early stages of
lesrning English. The number of students available for study was therefore
limited to those who met that condition. There were only four in the 1980-81
class, so I chose Andy from the following year. The students' first languages
are Burmese, Italian, and Korean. U Chal is a bit different from the other
four. Kis first and home language was Korean, but he had moved from Korea to
-7-
9
Brazil when he was five and spent seven years in Brazil before coming to the
United States. We will see effects of his Portuguese language background in his
writing.
(Figure 2 about here)
Eased on the teacher's ratings of the students and their scores on the
SurVey of Essential Skills Test administered near the end of their sixth-grade
school year, the students can be divided into three levels of English ability:
most proficient, U Chal; medium proficient, Michael and Andy; least proficient,
Laura and Su Kyong.
One interaction from each student's journal in the first week of February
is shown on the next page, displayed from the most to the least proficient in
English. There are striking differences in the five interactions, in terms of
topic, style, and language complexity, in both the students' and the teacher's
writing. For example, U Chal and the teacher are discussing dolphins, which
they were studying in social studies at that time, using relatively complex
vocabulary and structure. The teacher responds to Su Kyong's entry, which focu-
ses on more personal topics and is much more difficult to read, using short sen-
tences and simple language and vocabulary.
These texts provide an idea of some of the morphological errors that occur
in the writing (for example, missing past in "when I touch the skin," U Chal's
entry; missing plural in "you got more stamp," Michael's entry; missing article
in "I don't want dirty journal," Su Kyong's entry).
II Owls . . . I knew end I read that scientit was steadies Newt Welds.largest.. Lost year whoa I fuss la Brasil I was la ths beach sad I oowdophis dead ea the seed sad whom 1 tomb the skis le like east sad
thee When I est the lead smd I go to moo ths dolphis sees birds suseatlug the *aphis.
Tosehert The dolphins love eves bees tralmed to do madame work for the Navy.They seam to love ma istelligosQe. The birds help t sloes the beachby males dui deed animals. TM doleblo'r skis hos so miles maexpect ma amiss' that Books Ilka fish to hove stales. . . .
Michaels Tee !mow yesterday after wheal I give the Lisa may stamp amd Jemmysaid 'Mat did give her for sod I said Issue. she dos't hove thestamps es I have more me I most I. give bar the scamp mad yes got moresteep so I des't hove to give yaw the stamp.' Nrs. It if the' mark gmthe wall bow moeh they hove to pay the school. low Mew yesterday Igt to go the haspltal mod the mores said I hov, to drisk the boapall end I get the speed los sod se may II I have to go male.
Tesehart I's glad les west to the bospltal. It is importest that ye, tsks thebones of pills se yes will stay wall. Whom yes have takes all boxes,thee pow mey be all yell sod mot hove to take more pills. . .
.011111M11.
Amdy: 24ay I sm happy, because give um 'Now Jamul'. I like 'Paw Journal'.I Is picture mese Is 'Sores bird smd gores drape and sus'. I dmhappy. Ny mom tome here 5 day. I am happy. Ny mom give to so pro-em. "Nske ter. 'Stamps' Taves book (tartan) sod other is give tome. I sm happy happy. 'faster ammo hero mos please.' Today I smsewed esma to school. 'Oh, Del 0. I. Tomorrow I ms liret. 'Nomore'. ssly I ms hsppyl lappyz bappy3 happy,. Se yestomorrow. bps
Teaehart Isms Nether will be homy to see pal Tem ass tell yaw Nether yeaere lo. Iles to spook smd to write more logllah wiry isf
You ware st oebool before I mse today. Unreday we will go back sot -
sIde before school because Mrs. teed hos to go out to match all of thestmdests.
Ve learned shout the sell Aretle floss. ihuld yew like to live there?
Laura: tsar mre Used I like the red hart Is roily dee I hope yew Ilk, my.the ash use very good. Nrs. 4* sad I am plug Is i loyal I as mealyhappy. Oleos go to. I wish Ism a very happy Illestles to yaw msdpeer ODOM.
Teacher: I love year valestise. It Is so besotlful. The rod some sod the redboort ere se pretty. I like It os web because you made It for salflask you. It le sse of my favorite volestImes.
I hope your metes/ Is fun sad you come beet es TUesday ready to go tolevel GI
S. Yon: today I looks the Serval is del, it I do the sow Semi ead it thisjerssl is dely my bar le mot elys it I de the my mow jersol I Metmeat doly Wool sod today sisiag seadra Is mad ma emus shoo llamadoI hove to got the pljese le her heredity
Teachers Con you maks a big, big birthday oard for Ilasdra's birthday? Oho wouldIlke that.
I see your word study list Is poor journal. I sm happy you erestudyleg It.
If you hoop urItlag you will.get a sew jeursal.
Students' writing in the journals.
- -11BEST COPY AVAILABLE
Analysf.s
I chose to isamine the acquisition of six verb-related and Rout noun -
related morphemes in the journals:
Verb-related morphemes
Regular past Me played in the street.
Irregular past
Progressive -al
Progressive SE
Copula SE
Third person singular,present tense
Noun-related morphemes
Regular plural
Possessive -1
Definite article the
Indefinite article IL
These particular morphemes were chosen because whet. I read through the journals,
they stood out as part of the developing langusge compet-lice of these students.
Most were used sporadically or not at all initially, but their frequency
increased noticeably in the course of the year. One example of such change is
the difference in past tense marking in the following two narratives from
Nlchael's journal. In October he rarely marks past tense (once in this entry).
In the Nay entry all of the verbs referriog to pest actions are marked for tense
(verbs that should be marked for past are underlined).
Se sew the school.
Se is playing" in the street.
Se is playing in the street.
Se is a good student.
John likes school.
They are good students.
lie vent to Nary's house.
The teacher has a book.
Oct. 3
Mrs. Reed, Today I e to meth class Mts. 0- giLa test so I meet3 and 1/2 not finish becuss time is up. I mean she V.,* the 5.So I finish 2 -1777She Aka the math test is easy Niir-Seed.
May 14Yesterday I want home ay mother told me to study the Languarethat you AEI me to study I did study it
-10-
12
A. mentioned earlier, the acquisition of thess morphemes in the speech of first
and second lasguage learners has been well documented, providing a basis for
comparison with these written data.
The research questions that guided the analysis, designed to address the
issues raised above, are the following:
1. Are patterns of morpheme use in the journals similar to patterns found
in previous studies of morpheme acquisition?
2. Are patters. of acquisition smaller among the students?
3. Is there evidence of increased proficiency over time in the use of
these morphemes in the dialogue journal writing of beginning ESL students?
To address Question 01, I utilised methods used in previous morpheme stu-
dies that looked at suppliasce of morphemes in obligatory contexts. I divided
the ten months of writing into three sample periods --fall, minter, and spring--
of twenty interactions each (about four weeks' writing for each sample), and
analysed for each sample the propane@ of the selects: morphemes in the con'-xts
that the students themselves bad created in their own writing, treating each
context 'as a kind of test item Ihich the (student] passes by supplying the
required morpheme or fails by supplying sone ." (Brown, 1973). A morpheme
was considered supplied *ether or not ite form was correct. For example, is a
sentence like, 'They is goisg, the progressive auxiliary $E is considered
supplied.
Results of this asslysis are shown in Table 1. AwInimum of five contextsfor the occurrence of a given morpheme for each student was set. Where there
were fewer than five contexts, that morpheme use excluded iron the analysis.
Thus, in some cases there are gaps in the table --for example, for possessive,
in the winter sample, for U Chal, Laura, and Su Sloop This does sot seen that
so potential instances for use of the morphs's. were found, but that there were
fewer than five (use of the morphoses, Whether or sot there are five instances,
is discussed later, in the longitudinal analysis).
-11-
13
After each context for a morpheme was scored as: morpheme supplied (1 point)
or morpheme not supplied (0 points), the total number of times the morpheme was
supplied was divided by the total number of contexts in which it should have
been supplied, for an individual performance score, expressed as a percentage.
Scores for the five students as a stoup were summed in two ways --the Group Score
is the sum of the individual ratios and the Group Mean is the sum of the indivi-
dual percentages derived fron the ratios (following Dulay and Burt, 1974a,
except that to obtain a Group Score, they included all subjects in the calcula-
tion, even if there was only one obligatory occasion for the use of a particular
morpheme by that subject. Bore, a minimum limit of at least five obligatory
occasions for a morpheme to occur was imposed). You will notice that while the
definite article the and indefinite article a appear separately in the list of
morphemes to be studied (page 10), they are grouped in Table 1 under the
general category, Articles. This is because this grouping is done in the stu-
dies used for comparison. Me will look in more detail later at patterns for use
of articles.
(Table 1 about here)
Once group performance scores are obtained, they can be ranked from highest
to lowest, based either on the Group Score or the Group Mean. Since there is
little difference in rank orders by the two methods of calculation and since it
is the Group Score that is reported in most rank order studies, this score is
used to rank the morpheses here. This ranking is shown in Table 2.
(Table 2 here)
The rank orders are nearly identical in the winter and spring samples.
Only the past regular and third singular exchange ranks in the spring, and these
scores are based on very low numbers for each student, with no scores for sone
-12-
14
of the students on these morphemes. The sisilarity in rank orders among the
three simple periods is reflected in significant rank order correlation, as
shown in Table 3. In the fall there are several differences in the rank
ordering, but orders are still sinilar to those in the winter and spring, shown
by rank order correlation significant at the .05 level.
[Table 3 here)
Since an adequate number of instances of morpheme occurrence is
available for most of the students in the spring, this sample was chosen for
comparison with four studies of the sorpheme production of adult ESL learners--
in speech elicited by means of an instrument (the Bilingual Syntax )easure
(Bailey, Madden, and Krashen, 1974)); in free speech (Krashen, et al., 1977); in
compositions written quickly and not edited (Houck, Robertson, and Irashen,
1971); and in the same compositions, edited (Houck, et al., 1978) --and with
Krashen's 'natural order (treating the morphemes as if they ranked in linear
fashion). Table 4 shows the group rank orders along with group perforeance
scores in each of the studies.
(Table 4 Isere)
Rank order correlations between the results in each study and this study
show that group rank orders are imilar, as shown in Table 5.
(Table 5 here)
So far, we have found that rank orders for use of selected morphemes, when
results from the five students are grouped, are similar among the three sample
periods of this study. Rank orders found in this study are similar to those
found in others, using both spoken and written data.
Question 02 investigates *ether the rank orders are similar among the
individual students in this study. Following Andersen (1977, 1978), I used an
-13-
15
implicational scale to compare morpheme ranks for individual students both with
each other and with the group ranks (Table 6). The morphemes are ranked on the
left roughly according to the Gtoup Wean. Where a morpheme for a student
deviates from the group rank order, it is pAaced in parentheses. There are few
such deviations. The best scale follows the Group Wean rank orders almost
exactly, except that in the fall the ranks for articles and plural are switched,
and in the spring the ranks for past irregular and plural are switched (these
are marked with )). In both cases, the mean scores for the two morphemes are
very close.
[Table 6 here]
Although there are some deviations in each period, there is a great deal of
similarity between group sad individual ranks and among the individual students.
To demonstrate this, I calculated a "coefficient of reproducibility,"1 first
used by Guttman (1944) in variation studies and suggested by Andersen (1978) as
a way to investigate variation in language acquisition studies. The coefficient
of reproducibility for each implicational scale is as follows:
Fall 78.82; Winter 87.82; Spring 82.92.
Although lie, et al. (1975) consider 90 percent or above a valid scale, Guttman
points out that 85 percent is generally considered a sufficiently predictable
scale. Thus, the scale for the winter sample can be considered valid and the
spring sample is very close.
To summarise the analysis so far, the methods used in previous morpheme
studies have been employed here to determine rank orders for morpheme use in
required contexts for five students from four different language backgrounds --2
Koreans, 1 Burmese, 1 Italian, and 1 Korean with Portuguese as a second
language. We have seen that orders for these five students as a group and as
individuals are similar to orders found in those studies.
-14-
16
What has this analysis shown us? First, it has deaonstrated that a quan-
titative study of grammatical features can be conducted on dialogue journal
data, even with very beginning ESL learners. Before the analysis was begun, it
was not clear that the students' writing would allow such determinations. Early
in the year, the students' entries were short, handwriting was often difficult
to read, and passages were ambiguous in terms of contexts for morphemes. Later
in the year students wrote more, and the writing became much clearer and easier
to work with.
Second, methods of analysis siallar to those used la previous uorpheue
studies reveal patterns of morpheme use in dialogue journals siallar to those in
other relatively unmonitored productions, both spoken and written. That is,
certain morphemes - -progressive auxiliary, progressive 7LBL, and copula --are used
auch more frequently by most students where required then others like possessive
and third person singular -s and past regular -ed. This gives a rough indica-
tion of grammatical petterns that can be expected in the dialogue journal
writing of beginning ESL learners. If we were to stop the analysis here, we
would be left with the conclusion that these data confirm universal processes of
language acquisition, regardless of first language background.
However, the methodology used so far has provided a general starting point
for a more detailed longitudinalanalysis, in which a great deal of individual
variation becoaes evident. What follows is a discussion of acquisition patterns
in the journals of individual students over the three sample periods, in order
to address Question i3. In this part of the study all morphoses are included,
whether or not there are five contests for their occurrence. The first'thing
that is evident here is that the analytical method used greatly affects acquisi-
tion patterns that are found. Therefore, this discussion will focus not only on
patterns of change over time in the use of the morphoses, but also on setters of
methodology. Figure 3 shows the individual students' use of the four =un-
related morphoses, the, a, plural and possessive -'s, in required contexts in
-15-
17
the three sample periods. Patterns for use of Plural come from percentages
shown in Table 1. Tokens and percentages, not shown for the and a separately in
Table 1, are shown below in Table 7. Since instances of possessive -'s are
generally so few (except for Andy's journal), results are shown in Figure 3 as
fractions rather than as percentages.
Several patterns in these Charts are worth noting. Plurals start out high
in comparison with the other morphemes in the fall and then decrease in use for
most students, with only U Chal showing improvesent over time in the use of
plural -s. By the end of the year, their use is clearly not mastered by anyone
except possibly U Chal.
(Table 7 here(
(Figure 3 here(
The use of articles shows an interesting pattern. Although articles are
treated as a.single category in many of the morpheme studies, some researchers
who have separated then into definite and indefinite categories have found that
they desonstrate very different patterns of acquisition for second language
learners (Andersen, 1977; 'Ante, 1976; Rosansky, 1976). The patterns of
article use in the dialogue journals confirm these findings. In the earlier
analysis, in which articles are treated as a single category, they rank quite
high (after progressive auxiliary and lag and copula). When they are
separated into definite and indefinite categories, however, they demonstrate
very different patterns of use. Definite article the is used correctly con-
siderably sore frequently than indefinite a by all students, except for U Chal
and Laura in the spring, where they have mastered the use of both. Andersen
(1977), Dulay and lurt (1975), Makuta and Canciso (1977), Ressler and Liar
(1979), and Rosansky (1976) argue that patterns of article acquisition reflect
transfer fron the learner's native langusge. The dialogue journal data seen to
confirm thiso Laura and U Chal, whose previous languages (Italian and
-16-
18
Portuguese respectively) have articles, supply both the and a in obligatory con-
texts very quickly and reach 1002 accuracy on both by the end of the year (it is
important to emphasise previous, rather than first language here because, as will
be recalled, U Chal's first language is actually Korean and his second language,
Portuguese). However, Su Xyong, Andy, and Michael, whose previous languages
(Korean and Burmese) have no articles, never reach 902 accuracy. Use of a
reesins far behind the throughout the year, and Michael's and Andy's scores for
use of a decrease over tine.
Many second languageresearchers (Andersen, 1977; Hakuta, 1976; Lightbown,
Spada, and Wallace, 1978; and Pica, 1983, among others) argue that in the study
of morpheme acquisition it is not enough to consider only the suppliance of
e orphemes in obligatory contexts. For many learners, numerous instances can be
found of overgeneralisation to inappropriate contexts, and it is therefore
e islesding to state that a morphosis has been °acquired" uben its function bas
n ot been mastered. This is certainly the case with the use of articles in the
dialogue journals. Along with learning to use the correct article when one is
required, these students are also learniug not to use articles where none is
required. For Michael, Andy, and Su Kyong, overgeneralisation2 of articles is
almost as frequent as their omission, and in their journals we find sentences
like:
I have two sisters at tbe Bursa.
You know what happen to the some of the people.
I saw the many gase.
In order to determine each student's total control of the function of articles
(rather than simply use in obligatory contexts),Afollowed the method used by
Hakuta (1976) to calculate mastery of the use of a morpheme, the percentage of
tines it is used in appropriate, contexts. The nuleater of times the article
appears in correct contexts is divided by the total number of times it is used,
-17-
19
both correctly and overgeneralized to inappropriate contexts. This is shown in
Figure 4.
(Figure 4 here)
The development of appropriate article use is not at all uniform among
students and, like the use of articles in obligatory contexts, variation appears
to reflect first-language transfer. While U Chal and Laura show consistently
high scores for both definite and indefinite articles in appropriate contexts,
Su Kyong's, Michael's, and Andy's scores romain quite low, again with a lagging
behind the. Again, Michael's and Andy's appropriate use of a fluctuates over
tine.
Use of possessive -'s also seems to reflect transfer from the students'
first language. Contexts for its occurrence are very few in most of the jour-
nals, but from the little data available we can see that the two students who
supply -'s the most frequently are Su Kyong and Andy, the Korean students.
Korean has a possessive suffix on the possessor noun (Su Kyong-s chek
'Su Zpong's book"). Although this suffix is rarely used in casual conversation
in Korean, they have picked it up quickly in English. Andy not only Imes -'s la
obligatory contexts with high frequency, even writing things like, 'Today is
one's great great grandfather die day and, 'stick to one's own opinion,' but he
overgeneralises the form to many other situations, as if be is applying a rule
that whenever two nouns occur together, the first one must have -'s, as in the
following two maples:
Friday I am lend Sompob's money.
Today in the morning my kindergarten's sign isfinish.
&dusts (1976) found a pattern for possessive marking in the speech of his
Japanese subject, Oguisu, siallar to the one I fouod in Andy's journal. While
Uguisu marked plurals very infrequently, she reached OD percent accuracy with
-18-
20
the use of possessive -'s and overgeneralised the fors to possessive pronouns
(he's, they's) as well. Nakuta suggests that this could be a result of Japanese
influence - -in Japanese a postposed particle no appears in the same position as
the -'s.
Even though Michael frequently uses possessive noun phrases, he writes the
possessive suffix only once during the entire year. Frequently he uses long
possessive noun phrases, either omitting the -'s or using a possessive pronoun:
. my mother big sister son .
. . my father his brother wife
. . ay father brother wife her daughter . . .
Burmese has a possessive particle that follows the possessor noun (Maui* Ba
saou a Baum Bee book"). Since this particle constitutes separate syllable,
it may be that Micheal fails to notice the possessive suffix -'s in English. At
the same time, as two of the examples above indicate, hs may have transferred
the function of yeas a marker of possession to the English possessive pronoun.
Therefore, be might write something like, "Mum la her (for 71) book" rather
than, "Mount Ba's book."
In this longitudinal analysis of change ever time in the use of the
noun-related morphemes, we have seen no improvement in plural marking. The
articles the sod show very differeat patterns of acquisition, with correct use
of a lagging far behind use of the. Mere are also strong indications that
article use and overgeneralisation sad the use of possessive -'s reflect first
language transfer, a pattern that did not appear in the previous analysis. It
seems very clear that while certain trends can be identified across students,
individual variation duo to first language transfer can in no way be discounted.
Now we turn to the verb-related morphemes. Figure 5 shows the individual
students' use of these morphemes in required contexts in the three sample
periods.3 (The placement of progressive auxiliary and lag together, AUX and
-19-
21
INC on the chart, will be explained below.) Here patterns of acquisition among
individual students are more uniform. Copula is used nearly all of the ties
throughout the year by ell five students,4 while third singular -s and regular
past -ed are rarely used by any of then, and there is little or not inprovement
over tine. There is an increase over time in the use of irregular past in all
of the journals, and especially in U Chal's and Laura's, who reach over 902
accuracy in the spring. There Ls an increase in all of the journals except Su
Kyong's in the use of progressive auxiliary be and -13i with verbs in
appropriate contests.
(Figure 5 here)
Some interesting issues arise when we take a close look at forms used by
the students to express the progressive, which requires the grammatical morphemes
auxiliary be and :lsg. In tbe previous analysis it appeared that :lswas used
more frequently by all of the students than the auxiliary in contexts for the
progressive during the three sample periods (Table 6). This pattern is shown in
Figure 6 for the five students as a group. In this analysis, all BE + Verb
("She is go") constructions that were ambiguous as to shether or not they ware
progressive ware excluded, following the methodology used in most norpheme stu-
dies. Lightbown (1983), for examplc, 'counted as obligatory contexts for is
all obligatory contexts for the progressive, whether or not the auxiliary was
supplied...based on [the) assumption that added to the main verb is the
sore salient and essential sinker of the progressive..." (p. 226). Then she
counted "as obligatory contexts for the auxiliary ollzuCterances containing a
verb with laflection" (p. 226). Therefore, constructions such as "She's
blow the candle," in which bs but lot -13L is supplied, were not considered as
possible progressive constructions. There is good reason for this decision.
In such constructions, where there is no 1-is on the verb, it is often
-20-
22
impossible to determine whether the construction is an attempt at forming the
progressive or simply an overgeneralisation of be.
[Figure 6 here)
The problem with using this analytical approach with the dialogue journal
writing in this xtudy is that it obscures a lot of the data, in which there are
oany SE + Verb constructions without the zim inflection, as in the examples
shown below. Some (1-3) are clearly progressives; others (4 6 5) are clearly
not progressives. Most (6 6 7) are ambiguousthey could be either progressives
or simple present or pest tense constructions.
progressive
1. and so I told him I am o now.(. . I's going now.
2. I'm so to finish my homework.(I'm going to finish
3. today Sandra going to my home and got the Sandra.home.(I'm golig to Sandra's hose.)
Progressive not possible
4. Yesterday eight is telephone message the my grandmotheris die that is bsd message and sad message.(The rest of the context makes it clear that his grand-mother had died.)
5. today lunch time I. oo like lunch but I's hungry(I didn't like the-nisuEE:r-
ambiguous - progressive or staple present/pasty
6. Sunday raining and Monday is raisin but today is notraining but sun is come(The sun is cooliiraTtcame out.)
7. I am et la a diet because am fatt. Leticia is watching.vithing I eat and I hope I can be skinny because isgood.
(I am going on a diet/I am on a diet)
If all SE * Verb constructions that are clearly progressive (1-3 above) as
well as those that are ambiguous (6 6 7) are included in counts of use of
-21-
23
progressive auxiliary and :Am, a ouch different pattern appears from that shown
in Figure 6. In this pattern, shown in Figure 7, .-As. appears far less fre-
quently than the auxiliary in the fall, and slightly more frequently only in the
spring, an a priori decision to count as a context for the auxiliary 2.pa verbs
already marked with is inflection excludes from consideration the possibility
that some ESL learners may use the auxiliary to mark the progressive before they
use is. This appears to be the case in some of these students' dialogue jour-
nals (and Andersen, 1977, found a similar pattern in the written compositions of
ESL college students).
(Figure 7 here)
Andersen (1977) suggests an alternative method for analyzing the incipient
stages of progressive formation. Be argues that ratber thin treating auxiliary
be and is as separately occurring morphemes, as has been done in most morpheme
studies end so far in this one, it is sore informative to note the occurrence of
all of the possible forms used by ESL learners to express the progressive.
Figure 8 shows the results of this analysis in the dialogue journal data. Three
verb forme are used in contexts in which it is possible, that students are
expressing the progressive: AUX 4. Verb ING ("I essoing"), Verb 4. ING ("I
going"), and AUX + Verb ("I gue--this includes etbiguous AUX 4. Verb
constructions). (In the fall, U Chal also uses constructions with the verb
alone, such as "lilt° do." Since this construction is limited to U
in the fall only, it is not included in the figure.) The frequency of use of
each construction is expressed as a percentage. What we see here is that,
rather than using Verb 4. INC without AUX to express the progressive, Which Ls
the pattern found in soot morpheme studies and in the earlier analysis in this
this one, early in the year some students seem to use AUX 4. Verb instead. As
time passes, they learn not to use AUX 4. Verb alone and to use AUX Verb 4. LNG,
a process that was obscured earlier.
'Figure 8 here)
-22-
24
When we look at the progressive constructions of individual students, we
again find variation that seems to be due to first language transfer (overall
patterns of es* of AUX and ING by individual students are shown in Figure 5, but
cf. Xreeft 1984 for more extensive discussion of individual variation in the
formation of the progressive). Neither Laura nor U Chal, with Romance language
backgrounds. ever use the AUX + Verb construction in contextm for the
progressive, while Su Kyong, Andy, and Michael again have patterns similar to
each other --using AUX + Verb first, with AUX + Verb + ING appearing later.
Conclusions
I have taken two analytical approaches to the study of morpheme acquisition
of beginning ESL learners during eight months° time, as reflected in their
dialogue journal writing. In the first, I utilised methods employed in most
previous morphese studies, which are primarily cross-sectional, in order to com-
pare acquisition patterns in dialogue journal data with speech and other written
data. I found considerable uniformity in acquisition orders between this study
and others and among the individual students in this study, indicating universal
patterns of acquisition. In the second, I took a more longitudinal approach, to
look at patterns of change over time and to investigate which morphemes were
being acquired and Which ware not. In this sore detailed analysis --which
involved looking at overgeneralizations of morphemes as well as their use in
obligatory contexts, separating definite and indefinite articles rather than
grouping them under single category, and looking at progressive auxiliary and
-Is together rather than as separately occurring morphemes --I found interesting
patterns of acquisition, and while there were certainly trends common among stu-
dents, there was also a great deal of individual variation that was obscured in
the first approach.
Although the number of students involved la this study is adeittedly very
small, we can nonetheless draw some conclusions from these findings, make some
-23-
25
predictions about what we might expect in the dialogue journal writing of other
beginning ESL learners, and suggest acme isplications for classroom practice.
The first set of conclusions and implications has to do with patterns of
change over time in the use of grammatical morphology in the journals. These
students had very little trouble with the use of BE as copula, even at the
beginning of the year, and showed rapid mastery. For most of the students,
there were suostantial gain. in the correct use of the progressive auxiliary and
:Lim and the past tense marking of irregular verbs, atd U Chal and Laura reached
mestery on all of thee by the end of the year. At the same time, there was
little or no gain in tho past tense marking of regular verbs, and in the use of
plural and third singular -s. Only Andy showed considerable gain in the use of
possessive -'s. While U Chal and Laura mastered the use of the articles, the
other three studelts showed little isprovesent in their use.
Why these gains with some morphoses and not with others? Various reasons
have been proposed for morpheme acquisition orders in oral language, including
semantic and syntactic complexity (11rown, 1973; deVilliers deVilliers, 1973),
frequency of occurrence in the input to the learner (Larsen -Tresses, 1975, 1976;
Long, 1981; ?bark, 1980), and perceptual salience of the morpheme (Hakuta, 1976;
Labov, 1969; Robin, 1971). Iskuta argues that in speech "overtly marked"
forss, in mhich the inflection takes the form of a new syll.ble (such as irregu-
lar past, progressive:Au, prepositions and articles) are acquired earlier.
These forms "penetrate the attention of the learner. If the learner is moti-
vated to make his production match what is heard in the input, these forms are
the first to be acquired, because they are salient to the learner" (1976,
p. 336). Other forms with nonsyllabic markings, such as possessive, plural and
third person singular -s, and past regular are more difficult to decipher in the
input and less likely to be noticed.
The dialogue journal data sees to indicate that in writing, as in speech,
morphoses that ere syllabic (as are copula and progressive be when uncontracted;
-24-
26
progressive :As; and past irregular, as in "He went") tend to be acquired sore
quickly, sad therefore msy be more salient to the learner. Perceptual salience
is slippery coacept and is mot always tied to syllabicity. Copula and
progressive be are often coazracted and therefore not always syllabic. Andy
quickly picked up ea the use of progress 's, which is mot syllabic. However,
syllabicity does seen to be playing a role here.
When rules for the use of syllabic 'morphemes are easily learned (which is
the case for copula amd the progressive morphemes) or require the learning of a
sew word (with pest irregular verbs), it say be that they can be acquired la a
asturalistic, communicative context. (And Pica, 1982, 1983 found that explicit
lastructioe la progressive :1_4'6, en easily learsed rule, retuned in students'
sot may Issis, it cossisteetly la required contexts, but also overgeneralleing
it to coatexts la match it was mot required.) Mmen roles for syllabic aorpbeass
are difficult to learn (uhldb is the ease for articles), esturalistic acquisi -
ties, or at least their use la asturalistic communicatioa, does sot seem to
occur as ',sickly. Laura sad V Chal did master article use, but they could
transfer roles for article use le Italian sod POrtuguese to English. The other
students hod sore troeble glee articles.
Morphemes that tand mot to be syllabic (plural, la "two girls"; possessive,
la 'John's shoes"; regular pest, la "she mewed snort') are acquired more
slowly. It may be that these earphones will simply sot be used by ESL Isenberg
in commalcative, relatively unnoeitored writing contexts, even though their
rules ere easily tame'. and loomed, sad they have been taught and drilled
exteusively in class (as they had been in this class). Getting students to use
articles sal aoa-syllaloic morphemes correctly la thair wales my require
teach's( them methods for carefully 'WM% pieces they have writtea. Mere is
evidesce (cf. Pir, 1964) that direct lastructioe does accelerate the accurate
see of easy-to-leers morphemes ouch se third tiagular -e. Pica does not,
however, specify the coaditioft ender latch they are produced. What we have
2 7
seen in these data is that this knowledge, if in fact these particular students
display it in other writiog, has not yet been transferred to their dialogue
journal vritiog.
Second, while there are some overall trends common across students, there
is also considerable individual variation in their acquisition processes, which
seems to be a result of first language traosfer. Therefore, as Gess and
Selinker (1983) argue,
it is indeed possible and not incompatible to view second language
acquisition as both (1) a process of hypothesis testing in which
learners create koowledge from the second language data they have
available to them while at the same time viewing it as (2) a
process of utilisieg first language knowledge as well as knowledge
of other laoguages known to the learners in tbe creation of a
learner language. (Ir. 7)
Lasguage universals and uatural orders of morpheme acquisition not -
withstauding, we do mot want the search for end belief in universal processes to
blind us to the richoess of individual variation tbat is also present, to the
exteut that we develop laappropriato sod self-defeating empectations of our stu-
dents. Vs will sisply frustrate ourselves amd do our students a great disser-
vice if we aspect tbem all to perform alike.
Third, we Wive found that dialogue jourmal writing does reflect Changes in
students' lagguage proficieecy over time, even at the earliest stages of secood
language acquisitios. Vs were able to quantify and plot this grontb for each
student, includieg the least proficient in English. Although this particular
study does sot investigate this issue, it way be that the dialogue journal
writieg not ooly reflects but also facilitates that growth. This possibility is
certainly supported by 0! qualities of the writing, those considered optisal
for first sad secoed language acquisition, as outlined in the beginning of this
chapter.
-26-
28
It is clear that dialogue journal writing can serve as a valuable resource
for teachers, enabling them to follow their students' progress in extended
writing that is student generated, written for purposes other than evaluation,
and relatively unmonitored. Frou reading student entries, teachers can find out
a great deal not oely about their students as human beings, but also about what
they are learning, where they might be having trouble, and where future lessons
might focus. Although we have looked only at the acquisition of morphology in
this study, the writieg yields a wealth of information about each student's
progress, from the smallest features of the language all the way up to discourse
and interactioe patterns.
Finally, from the various esanples of text shown throughout the chapter,
it is clear that ISL learsers can compose and express themselves in teglish long
before they love metered its foras and structures. Ives at the beginsing of
the year, Oben morphology was st the earliest stages of development, these
'Modesto more homy writing Bed reading meaningful text in their journals. As
the year progressed they wrote narratives, described events and problems, and
argued their points of view. In the meantime, language forms and structures
contieued to develop.
-27-
29
Notes
1
The formula for calculating a coefficient of reproducibility is:
no. of deviationsR 1 - ((no. of rows)(no. of columns)) - no. of empty cells
7Yor the spring sanple, the formula would be: R 1 - (9z5) -4 .829
2Pica (1983a) and others use the tern 'overapplicatione to cover both
what they call "over-uses" (use of a morpheme where no morpheme is required) and
"overgeneralisations" (use of a morpheme share a different morpheme is
required). However, since the term "overgeneralisatione has traditionally been
used to cover both situatious, I will continue this use of the term here.
3Patterns for use of copula, past irregular, past regular, and third
singular -s come from Table 1. Where there are so percentages in Table I frac-
tioes are given in Figure 5.
4There are also many instances of overgeneralisations of be to
Inappropriate contests. Overgeneralisations of all of the morphemes in this
study are discussed in &soft, 1964.
Acknowledsesenta
The research for this study ems conducted with the support of the National
Institute of Education, NIE-G-83-0030. I am grateful to the National Institute
of Education and the Center for Language Education and Research for continuing
support of research on dialogue journal writing. I am also grateful to Ralph
Fasold for performing the statistical analyses and to Ralph Fasold, Roger W.
Shuy, Jona Staton, and Walt Wolfram for useful comments on drafts of this
chapter.
-28-
30
References
Aodersen, R. V. (1977). The impoverished state of cross-sectional morphemeacquisition/accuracy methodology. In L. A. Henning (14.), Proceedings of theLos Angeles Second Language Acquisition Research Torun. Los Angeles, CA:UCLA. (ERIC Document leproduction Service No. ED 146 780)
Andersen, R. V. (1978). An implicatienal model for second language research.Language Learning, 28, 221-282.
Bailey, N., Madden, C., 6 Krashen, S. D. (1974). Is there a "natural sequence'in adult second language learning? Language Learning, 24, 235-243.
Brown, R. (1973). A first language: The early stages. Cambridge, Mk:Harvard University Press.
de Villiers, J. G., 6 de Villiers, P. A. (1973). A cross-sectional study ofthe acquisition of grammatical morphemes in child speech. Journal ofPsycholiiguistic Research, 2, 267278.
Daley, R. C., A Surto M. 1. (1972). Goofing: An indicator of children'ssecood language learning strategies. hilmage Learninl, 22, 235-252.
Daley, N. C., 6 Burt, M. 1. (1973). Should um teach children syntax?Language Learning, 23, 245-258.
Dulay, H. C., 6 Burt, M. 14 (1974a). Natural sequences in Child secondlanguage acquisition. Language Learning, 24, 37-53.
Dulay, N. C., 6 Burt, M. 14 (1974b). A new perspective on the creativeconstruction process in child second language acquisition. LanguageLearning., 24, 253-278.
Delay, 16 C., 6 Burt, M. H. (1975). Creative coastruction in second languagelearning and teaching. In M. 1. Hurt 6 H. C. Dulay (Eds.), New directionsin mimed language learning, teachlago and bilingual education (pp. 21-32).Vashimgton, DC: TISOL.
Gass, S., 6 Selinirsr, L. (1983). Introduction. In S. Gass 6 L. Santa:srEds.), Language rrarefer in language learning (pp. 1-18). Rowley, NA:Newbury Nouse.
Nakuta, 14 (1976). Becoming bilingual: A case study of a Japanese childlearning Inglish as a woad language. Language Learning, 26, 321-351.
Nahuta, K., I Concise, N. (1977). Trends in second language acquisitionresearch. Harvard Educational Review, 47, 294-316.
Vouch, N., Robertson, J., 6 Rrashen, S. D. (1978). On the dosain of theconscious grammar: Morpheme orders for corrected and uncorrected ESL stu-dent transcriptions. IISOL Quarterly, 12, 335-339.
Kessler, C., 6 Idar, I. (1979). Acquisition of English by a Vietnamese motherand child. Working Papers on 111151=110111, 18.
-29-
31
Erashen, S. D. (1977). Soso issues relating to the monitor model. InR. D. Brown, C. lorio, 4 R. Cry... (Eds.), Teaching and learning Englishas a second language: Trends in research and practice (pp. 144-158).Washington, DC: TESOL.
Krashen, S. D. (1982). Principles and practice in second lansuase acquisition.Oxford, gngland: Pergason Press.
Erashen, S. D. (1984). Writing: Research, theory, and applications. Oxford,England: Pergamon Press.
Rrashen, S. D., Houck, N., Giunchi, P., Bode, S., limbs's', R., 6 Strei, G.(1977). Difficulty order for grammatical morphemes for adult second languageperformers using free speech. TESOL Quarterly, 11, 338-341.
Kreeft, J. (1984). Dialogue journal writing and the acquisition of granmaticalmorphology in English as a second language (Doctoral dissertation,Georgetown University.
Ireeft Pinola, J. (1986). Literacy through written interaction. Passage: AJournal of Refugee Education, 2, 24-29.
Ereeft, J., any, 1. V., Statea, 3., Seed, L., 4 Worroy, 1. (1984). Dialoguegratin: Analysis of student-teacher interactive writing in the learnintof English as a second lamosse (Report No. 11.1-043 -0030). Washington, DC:Center for Applied lieguistics.
Labov, V. (1969). Contraction, deletion, and the inherent variability of theEnglish copula. Language, 45, 715-762.
Larsen-Freeman, 0. (1975). The acquisition of grammatical morphemes by adultESL students. TZSOL Quarterly, 11., 409-419.
Larsen-Freeman, D. (1976). An explanatioa for the morpheme acquisition orderof second language learners. Language Learning, 26, 125-134.
Lightbovn, P. N. (1983). 11xplorimg relationships between developmental andinstructional sequences la L2 acquisition. la N. V. Seliger 4 N. N. Long(1da.), Classroom oriented research in second language acquisition,(pp. 217-243). Rowley, NAa Newbury Nouse.
Lightbown, P. MI., Spada, N., I Wallace, 1. (1978). Some effects of instructionon child and adolescent ISL learners. In 1. C. Scarcella 4 S. D. Irashen(Eds.), Research in second language acquisition (pp. 162-172). Rowley, MA:Newbury Mouse.
Long, M. N. (1981). Questions in foreigner talk discourse. Language Learning,31, 135-157.
Pica, T. (1982). Second language acquisition la different language contexts(Doctoral dissertation, University of Pennsylvania).
Pica, T. (1983). Adult acquisition of English as a second language underdifferent conditions of exposure. Language Learning, 33, 465-497.
-30-
32
Pica, T. (1984). Ll transfer and L2 cosplexity as factors in syllabus design.TESOL Quarterly, 18, 698-704.
Nosansky, E. J. (1976). Second language acquisition research: A questionof methods (Doctoral dissertation, Harvard University).
Slobin, D. I. (1971). Developsental psycholinguistics. In V. 0. Dingwall(Ed.), A survey of limguistic science (pp. 279-410). College Park, KD:University of Msrylast.
Staton, J. (1980). Writing and =tinseling: Using a dialogue journal.Language Arts, 57, 514-518.
Staton, J. 1984. Dialogue journals as a means of enabling written languageacquisition. In J. Ereeft Peyton, 1. V. 'buy, J. Staton, 1.. Reed, 6I. Morroy, Dialogue writimg: Analysis of student-teacher interactive
(Report No..tasasecomwritiinthelreadlause
Washington, DC: Cmnter for Applied Linguistics.
Staton, J., Shay, R. V., Ermat Peyton, J., 6 Need, L. (in prase). Dialoguejournal communication: Classroom, linguistic, social, and cognitiveviews. Norwood, NJ: ales.
-31-
3 3
Length of Time in U.S.Country of First/Bome Schools at Beginning
Student Birth Language of Journal School Tear
1980-81 Classroom:
Michael Burma Burmese 8 months(arrived la U.S. 1-80)
Laura Italy Italian 5 months (arrived 440)
SU Iyoog Korea Korean 0 months (arrived 10-80)
U Qial Korea Korean 6 Booths (arrived 3-80)
1981-82 Classroom:
Andy Korea Korean 3 meths (arrived 5-81)
Figure 2. Students In the study.
if
aniargi
ft.....
ista4goold41
.....,..,
ii4isisie:
1i mtl-tRoo
sktie04e
ir t.
ez4Iv-IR
x=24aRaSa
00000
iid4iikA0
c lik
latgl=ms
ElKielf
4
itftet""te!,
Itittirre
Eiiiiitt
iatkii24
IRII4t-0
t 8244 10
f mr.,
it
isliumSsa
renigins:
OlgEsE
1xii*xmalfs
.;za:Mtier
:samiexiia
,
RIP.RmtrR!4!
assisiox;
!PM
tZ
g...4.:...
owmt1.4.41.
litimid4
1
:Riling
..0.0.......
ChuipAi
-gm
Main
iTod:Rto:
RVftSR*11°
pti.lxIsi
oo!Rt Ro
lit*,%1
in-is:tab
"11110
litasollis
Wall
R Rt
IllAellil
1:Ifilal
a/Gill2;
Mast:
...m.
1/11//eifti
te*Ett
2114atttitta.1
11:40:R^,
411,
ielt liSR1 R
sVeVnt
RD
=I=timsax
...,.r .7.
/140i0l2i
weee
ID
exassts%
/WWII;
"EW
Im.
?AutsaIlk
i110
II millin
m110
M mit113
m1
IX
M mid !I
Table 1.
Individual and
group accuracy percentagesan t no morphemes.
Fall Winter Spring
cop ing logarts cop copplu aux auxlog arts artsaux plu plupast irr past irr pest irrpest rag possess possess3 sing pest Mg 3 singpossess 3 sing pest Tag
Table 2. Group rank orders for morpheme suppliance by Group Score.
Correlation coefficient Significance level
Fall and Winter 0.75p (.05
Fall and Spring 0.77 p <.05
Winter and Spring 0.98 p (.001
Table 3. Spearman rho correlations between morpheme ranks for threetines during tbe year.
39
894-elicitedl Pres apeech2
Uncorrected
transcripts5
Co7rected
transcripts5
"Natural
ordee4
Dialogue
journals
cop $4
ins 83.7
plu 79
art 79
aux 69
Past itt 54
3 sins 41
cop 87
lug 84
plu 71
art 69
past irr 67
past res 64
aux 56
3 sins 36
ins 97
cop 97
past irr 84
aux 82
art 82
possess 75
plu 75
past rag 61
3 sins 60
cop 98
ins 97
past irr 87
aux 86
art $3
possess 80
plu SO
3 sins 76
pant res ol
ins 100
Cop 97.2
aux 82.2
art 77.4
plu 60.6
past irr 55.5
possess 45.2
3 sins 13.0
past rag 9.8
ins
plu
co
aux
art
past irr
past res
3 sins
possess
1. Bailey, Madden & &rashes, 1974. Imported in &rashes et al., 1977. p. 340.
2. Krasheo, Mouck, Gluschl, Sods, Birshaus 6 litrel, 1977, p. 340.
3. lieu& Robertson 6 &realms, 1978, p. 337.
4. Krashen, 1977, p. 149.
Table 4. Morpheus rank orJers in oral and written productions.
40
Rant orderStud5 llititatios Heat.' correlttion
Bailey, et l. 1974 1811 0.82b
[rashes, et al. 1977 Prat speech 0.68c
trashes, et al. 1,78 Ovcorrected transcripts 0.890
Ilasbes, et al. 1,78 Corrocted trsascripts 0.878
Ilasbes's
Natural order's 0.83a
a. p<.01, vr9
b. pC.05, s-7
C. p.06, es4
Table S. Spearnau rbo correlations of dialogue jourual gritting vitbotber studies of oral and grimes productions.
Mean U Chal Laura Su iyonA Michael Andy
Fall
91.2 100 90.0 73.9 94.4 97.8coparts 67.1!] 86.8 34.8 65.0 84.3 68.6plu 69.1 64.3 (90.9) 50.0 64.7 (76.5)log 60.2 8.3 -- (92.3) (80.0) --aux 43.3 8.3 -- (61.5) 60.0 --pest irr 17.8 (20.0) -- 16.7 25.4 9.13 slog 10.8 0.0 -- -- 21.4 --post reg 10.4 (20.0) -- 0.0 21.4 0.0possess 0.0 -- -- 0.0 0.0
Mister
99.1 100 100 100 100 95.5lagcop 92.9 100 (96.0) 80.8 92.2 95.3ass 81.2 94.1 100 65.5 92.3 54.5arts 72.2 02.6 100 56.4 91.7 20.4pse irr 521 75.0 88.2 38.9 51.1 10.7plu 34.2 65.5 64.3 (50.0) 38.6 (52.5)possess 32.6 8.0 (57.1)pest rag 18.4 40.0 40.0 0.0 4.2 --3 sift 5.7 14.3 4.2 0.0 (7.1) 5.0
Sprist
Log 100 100 100 100 100 100cop 97.5 100 100 94.3 95.7 97.4Oft. 00.0 (00.5) 100 80.0 76.2 45.2au 72.1 100 100 40.0 (83.9) 36.4post Ifs 62.7 Igg 94.1 32.0 64.3 23.3plu 60.3 80.8 71.1 (50.0) 44.2 (55.6)possess 41.4 50.0 33.3 -- 0.0 (82.4)3 sift 13.6 30.0 21.4 -- (3.6) 0.0post res 11.9 18.2 -- -- (17.6) 0.0
Table 6. Isplicatlossl scales for sorpheue accuracy orders. lased opthe army seam.
finite Article Fall Milner S rin
0 Cbal 50/57 87.7 69/75 92.0 51/52 98.1LIWATII 6/10 60.0 40/40 100 35/35 100
Su Kyeag 8/12 66.7 21/31 66.7 22/27 81.5Michael 39/43 90.7 120/130 92.3 102/129 79.1Altdy 14/22 63.6 18/71 25.4 26/48 54.2
TOTAL 117/144 81.3 268/347 77.2 236/291 81.1
Iadefinite Articl
0 Chia 4/11 36.3 17/20 85.0 13/13 100Laura 0/13 0.0 16/17 94.1 14/14 100So [yeas 0/8 .0 0/8 0.0 3/8 37.5Michael 3/8 37.5 0/14 0.0 16/43 37.2Andy 7/14 50.0 2/32 6.3 7/25 28.0
TOTAL 11/54 20.4 35/91 38.5 53 103 51.5
Table 7. Ose of definite and Indefinite articles in obligatorycontexts.
U Chal
100
90
80
70
60
the
a
Plural
SO
3/6
Possess
40
30
2/4
1/320
10
0
Lours
100
90
80
70
60
30
40
30
20
10
0
fall Neter Sprint
Figure 3. Change over time in suppliance of noun-related morphemes
in obligatory contexts. Individual students.
U Chal
100
90.994.4
98.1
the
90
80
80.086.787.3
70
60
50
40
30
20
10
0
Fall Winter Spring
Lours
100
90 5.7
ICI
70
GO
30
40
30
20
10
0
10095.2 a*
87.5418.9 the
Fall Winter Spring
There were no contexts for the use of a for Laura in the fall andSuKyong in the winter samples.
Figure 4. Appropriate use of definite and indefinite articles.
46
iMer_st100
op
so
ve
ms
0 33.342.0 ge
30
40n. 2,
30
30
300.0
0
iPe443
1sSi100
fo00 77.0
TO
00.0
10
M
30
V10
0 'err.-felt
Figure 4, cont.
1,4h)14-
M.7
.4-
CAMIAP
47
Chal
100
SO
$O
70
60
50
40
30
20
10
0
COP11011,ING
PAST ix,
3 SING
PAST REG
Laura
100
SO
SO
70
60
SO
40
30
20 1/4
10
0
Fall Slater Spring
4/10
COP
60X,INGFAST IRA
1/4
PAST RtG3 SLAG
Fall Winter Spriag
There are not contexts for use of third singular -s in Sultyong'sjournal in the fall and minter samples.
Figure 5. Change over time in suppliance of verb-related morphemes.
100
90
SO
70
W
50
40
30
20
10
0Ulster SprIsi
Figure A. Progressive morphemes, group scores. Ambiguous coustructioss secluded.
50
100
90
SO
70 AUX67.6
60
so
ao
30 35.3
20
10
0
84.6
79.5
91.1
113.0
Toll Mister Spriss
DIGAUX
figure 7. Progressive oorpbeles, grcup mores. Ambiguous coostructioss included.
100
90
80
70
6048.5
50
40
3019.1
20
10 16.2
0
65.4
19.2
AUX + Verb74.1 LNG
I se soleNt.
17.0 Verb ING
Irina.15.4
8.9 AUX 9erbI am so.
Tall Winter Spring
The percentages la the fall samples do sot equal 1002. Mere Chal'suse of the verb aloes (Is! to do...) mode up 162 of the progressivecoestructioes.
Tigers S. Coestructioss used to sepressive the progressive.